Hi Dale! Did some more tinkering today. I agree that other pairs besides the EUR/USD should be examined, but my understanding is that all pair movements are fractal in nature. Assuming market efficiency hypothesis is valid, it shouldn’t matter which pair you trade. The extra bucks that could be made in a more-volatile exotic pair like the AUD/JPY will be eaten by higher spread. If there was indeed an edge to be found, I think arbitrage trading will ensure that edge gets eliminated. Perhaps my thinking is wrong, though.
Anyway, performed two more backtests for the months of Oct and Nov 2010 on the M30 TF with TP=0.5SL and TP=0.33SL. In October TP=0.5SL performed slightly better than TP=0.33SL (349 pips vs 336 pips), and in November, TP=0.5SL ended up with a nominal profit of 568 pips versus 426 pips for TP=0.33SL.
TP=0.5SL performed better than 0.33SL for both months and it wasn’t what I predicted. I think at this point I’ll begin to clean up my results and tabulate them for people to see. After testing nine months worth of M30 data, however, I have little doubt that this system is profitable, at least where TP=0.5SL.
You’re either not QUITE as ‘new’ at this as you make out OR you have some or the other relevant (mathematical or scientific) qualification (what with talking about ‘the fractal nature of markets’ and ‘arbitrage trading’)??? LOL!!!
For what it’s worth: Bill Williams has written three books on the subject of the fractal nature of markets’!!! LOL!!! (Well: there’s three books but the information insofar as the fractal nature of markets is concerned is pretty much the same in all three). Bill Williams: ‘Trading Chaos’ or ‘The Profitunity Trading System’. It’s odd this. My FIRST book that I bought was Wilder’s ‘New Concepts In Technical Trading Systems’ and my second book was Bill Williams’ ‘Trading Chaos - Second Edition’ (and if memory serves me correctly I ordered them both from Amazon at the same time). Eventually I went the Wilder route (after the Williams route cost me quite a few USD)!!! LOL!!! Not only that: but believe it or not I’m one of those idiots that believes the more complex the trading system and the calculations the better it must be (even although Wilder HIMSELF would disagree with me i.e. by his own admission his best ‘offering’ is, or was, his Volatility System which is probably the easiest system ‘in the book’ to trade). Hence my fascination with Wilder’s Swing Index System!!! LOL!!!
Like I said: just don’t ‘overtinker’!!! We don’t want you ending up in some mental asylum (believe me: I got close at one point)!!! LOL!!!
Well, I didn’t get the spread eurusd comment this way. Just what my opinion is about this pair. It’s also farther away from manipulation because of the huge trading size imho.
The thing with too high or too low wouldn’t be an issue with a backtested system. As you know or maybe not(?) I prefer to backtest all of my methods for some years. That way I can play with different settings of risk/reward to find something what give me a good enouth edge. So, it was just an idea or two, as I thought about the sar as well and use it in one of my older strats, albeit not as stop, lol. I mean, it’s always the thing between too soon or too late or too high or too low. One must find a strat which is just about right with the entry/exit regarding statistics, imho. Well, you probably know all this, but the op might get new ideas.
Have a great day there and lots of $$$ (no pips, lol)!
It appears that a new dot appears simultaneously with the [U]open[/U] of a new candle. If price retraces, can that dot also change directions? If so, do you predicate your entries at the [U]close[/U] of the candle where an opposite direction dot appears? :33:
If the spot price retraces and hits the dot, the dot will flip over (change direction) in the same candle. So if a new candle opens and the dot appears at the bottom and the spot price falls and hits it, the dot will flip over to the top. Your SL will be triggered when this happens. At this same point in time, you can open a new position with your SL being the new dot, and TP=0.5SL. In other words, your entry will be wherever the old dot was when it was hit by the spot price. There would be no need to wait for the candle to close since the dot will flip over instantaneously. At least that’s been my experience with using Metatrader. I hope I’ve interpreted your question correctly.
Well I’m nearing the completion of my commerce degree at university so that may have something to do with it, lol. Before that I had an engineering qualification so tech analysis seems right up my alley. I’ve worked in the financial sector in the last year and absolutely HATED it. Meeting deadlines week after week made every day at work a battle. It got to the point where I couldn’t take it anymore and quit. I much prefer to use my knowledge for MY benefit rather than someone else, hehe. Enter forex. Everyone I spoke to about forex dismissed it, but I don’t think there’s much difference between daytrading forex and stocks. Longer-term, there’s probably an argument for stocks in terms of value investing, but I think forex is a better option for daytrades. For one thing, transaction costs tend to be cheaper. But I’m still learning.
With regards to this particular strategy that I’ve developed, I just want to iterate it’s very much a WIP (work in progress). With nine months worth of backtesting, however, the results seem solid. The only two dangers at this point is if my backtesting method is fundamentally flawed, or if live trading conditions siginificantly differ from backtesting (possible). I’ll be doing another backtestfor one more month and then tabulate my results.
Wilder is (was) ALSO an Engineer before becoming a trader!!! LOL!!!
Oddly enough: somebody noted somewhere around here that it would seem that all the famous (equity and commodity traders anyway) either had an Engineering Degree or a Degree in Psychology!!! LOL!!! And isn’t it strange that both are ‘inextricably linked’ when trading (technical analysis and trader psychology)!!! I’ve never thought about that until now. What’s even MORE strange: most of the most successful INDIVIDUAL traders in the world are Floor or Pit Traders (those that trade for themselves) and, so far as I can tell, these guys LITERALLY ‘come off the street’ but they’re ‘sharp’!!! LOL!!! Go figure!!!
Stocks and FOREX??? I’m not even going to TRY to have that debate with you (believe me: it’s only going to get ME into trouble once AGAIN)!!! LOL!!! Just remember: ‘you get what you pay for’ (referring to transaction costs)!!! LOL!!!
I’ll tell you this though: I’m keen to see what you end up with.
I’m depositing $2k this week and will keep a live trade journal to see how this “strategy” performs under live conditions. Wish me luck!!!
Edit: some caveats. The results above are for the M30 timeframe. It is best to avoid trading just before or after the weekend to reduce your risk. In terms of trend filters, skipping a trade after a loss is your “trend filter”. Since PSAR will fail in a ranging market, you should avoid being in one. However, it’s hard to identify when a trending market ends and a ranging market begins (at least for me) until you’re in the middle of one. Skipping a trade after a loss will reduce your exposure in a ranging market. The alternative would be to trade outside sessions that typically range (e.g. the Asian session). I may study this in future but my reasoning is that trades will often overlap different sessions on the M30 timeframe so beware.
Well: you’ve got my ‘vote’ and I’m ‘rooting for you’ for sure.
I know this is ‘rich’ coming from me (given what I was saying about the exotic pairs) but it’s a PITY about the ‘Spread Cost’. Then again: it’s probably the closest to any type of ‘symbiotic relationship’ I’ve seen i.e. the relationship between you and your broker. You ALMOST make the same (and you’re slightly ‘up’)!!! That’s fair. Actually: your ‘Spread Cost’ probably won’t be so high anyway so you’ll probably be ‘up’ far more I reckon (unless of course you’re trading with a broker with a fixed spread of 2 pips obviously).
Hmm, a backtest with drawdown. What a seldom happening, lol. Anyways, if I’d allow to suggest something: Don’t curve fit it into 10 months. Do it for not less than 2 years. One year in fx is typic very different to another one. If you make a backtest over two and your equity curve looks like a straight uptrend, then, and only then use it for trading. Not to say more years would be even better for testing, but imho 2 years are good enough if you look at all the other statistics and they look good. Plus the strategy is simple and has a little logic inside, lol.
It IS rather ‘unique’ and ‘refreshing’ is it not??? LOL!!!
Well whatever the outcome: it’s been nicely presented and well thought out and I hope it works out. I have to share a ‘little secret’: for the first time in years I’ve ‘lobbed’ Parabolic SAR onto my charts and been monitoring it on the 1-hour chart of the Dow (using this system i.e. I THINK we can now call it a SYSTEM instead of just a ‘system’???) and oddly enough the TP seems to get hit most of the time even when the Dow is trading in a range!!! Alright: monitoring a few hours worth ‘doth not a profitable trading system make’ but I think there’s a good principle here. It’s not ‘greedy’ and as long as one can live with missing out on POSSIBLE long ‘straight up’ or ‘straight down’ trends without wanting to ‘slit your wrists’ because you’ve only taken a fraction of what a particular trade COULD have made then I reckon you’d be ‘good to go’.
Yes, it is very uncommon that somebody is testing what he is doing before blowing at least one account, lol. Some even blow several accounts and give up before testing anything, mwahaha. Well, them are probably gamblers and have fun with it. Nothing wrong with that. Like Las Vegas. Nothing wrong with that too. However, some prefer to “sit” at the other end of the table, lol. A few.
Back to sar, I am pretty sure one can use it for a profitable system. While my last system is doing everything without any indicator I guess. Like the ole turtles did. Well, one exception: I have an indicator for my emergency exit rule. Not a sar, tho. I use a “seismograph” of price. If the pricequake is too heavy, I get out.
Yes, it is very uncommon that somebody is testing what he is doing before blowing at least one account, lol. Some even blow several accounts and give up before testing anything, mwahaha. Well, them are probably gamblers and have fun with it. Nothing wrong with that. Like Las Vegas. Nothing wrong with that too. However, some prefer to “sit” at the other end of the table, lol. A few.
I’ve never called YOU names!!! LOL!!! The above is my ‘style’ man!!! LOL!!! ‘Bet the farm’ and THEN when you’ve lost it all: find the faults in the system!!! LOL!!! Then: BORROW another ‘farm’, rinse, and repeat!!! LOL!!!
Yes but only when I ran out of ‘farms’ to borrow!!! LOL!!!
Bank of America??? Have faith!!! LOL!!! Don’t worry: I don’t own anything (didn’t buy). That means it’s almost GUARANTEED to double in price by 31 December 2011!!! LOL!!!
No, I don’t have faith. Particular with investments. I indeed lost it, together with some hard earned cash in my first stock investments. LOL!
No seriously, I am doing better in commodities. A company can lose all of it’s value, but a commodity can’t. Maybe sometime in the future Coca Cola, Google and IBM will lose all it’s value, but cocoa, sugar, yes even copper will always have some value.
Fantastic thread! Been playing with SAR for ages but refused to buy yet another book, including the ones suggested here. Got so many stacked up next to my bed it would be years before I get through them all. Now if only I could the Jonny Nemonic thing and stick a usb into my head and download all the knowledge. ahem!
So I have learned a lot from this thread. Some awesome ideas and thoughts. Will be testing this from now on a demo account.
1 thing i do want to clear up re Entry rules.
AS SOON AS I SEE A NEW DOT ON THE OTHER SIDE OF THE PRICE ACTION BEGIN TO FORM, I place an Entry order for the opposite direction at the previous CLOSED CANDLE dot? If this is true, I think it is that one thing I have been missing from my testing. I was waiting till the candle closed and then setting an Entry at current price +20 (-20) to catch the trend but often it would go up 22 and then drop 60.
Have thought about identifying the time placed against each open of each trade to see if certain times of certain days produce more of the lossing trades. You mention the avoiding the start and end of the week but what about certain times?
Rightio. My broker (or is it MT4) will only let me view data from the last 14 months though. But I also have other ideas to test as well, so I’ll see what other data sources I can find.
Edit: The actual backtesting was conducted over the months of Aug to Dec 2010, May to July 2011 and Oct to November 2011 for anyone interested. August 2010 is about as far back as I can go.
Yes, if I understand your question correctly. You open a position as soon as the old dot is hit by the spot price. This will flip the dot to the other side of the candle.
I’m pretty sure that there is an “optimal” time to trade with this strategy. I tried checking the trade sessions for one month but discovered that many trades will overlap different sessions so discovering the “optimal” times will be messy. It’ll probably be a fair assumption to think that the London and NY trade sessions would be the best to trade but I can’t confirm that yet. I originally formulated this strategy to see if it could function at a brain-dead level but in the coming weeks I hope to test different things out.
And I just want to add, if anyone is entering the market fresh after a period of absence, wait out the current “trade” and check if it results in a win or a loss. If it’s a loss, then skip the next trade. That way you don’t enter the market blindly.
Everyone I spoke to about forex dismissed it, but I don’t think there’s much difference between daytrading forex and stocks. Longer-term, there’s probably an argument for stocks in terms of value investing, but I think forex is a better option for daytrades. For one thing, transaction costs tend to be cheaper. But I’m still learning.
But I also have other ideas to test as well, so I’ll see what other data sources I can find.